concurrent function deployment

72 downloads 0 Views 1MB Size Report
WHATs normally define a set of end-user requirements about what a ..... Most commonly, a PDT team, through a row of feasibility matrix, establishes a set of.
CHAPTER

1

CONCURRENT FUNCTION DEPLOYMENT

1.0

INTRODUCTION While manufacturing philosophies have changed drastically during the eighties, the pace of such transitions from concept to practice has been very slovv. Despite painful restructtiring, reorganization. and even process re-engineering efforts, both the European and U.S. automotive industries have failed to attain parity in product cost, productivity, or throughput with Japanese producers and transplant operations. Earlier published work sho\ved assurances that the co1npetitive gaps could be closed using Quality Function Deploy1nent (QFD) or siinilar programs. This had n1otivated abandonn1ent of n1any traditional function values in favor of employee empo\verment and autono1nous multifunctional tean1work. Many such combinations have been tried with QFD, along with Pugh's concept [Clausing, 1994], voice of the customer (VOC), and product development team (PDT). They are discussed in volume I of CE Fundamentals. Though each QfD con1bination provided new opportunities and contributions toward cost and productivity iinprovements, such programs have encountered difficulties in making a parent company globally competitive. Furthern1ore. the gains that \vould seen1 obvious and feasible through the exploitation of QFD and its combination (in a quantifiable con1petitive sense) have not always been fully realized. The application of QFD is a fairly old (over two decades) idea [Hauser and Clausing, 1988]. Historically, the concept of QFD was introduced by Japanese [e.g., Mizuno and Akao, 1978; and As wad, 1989] in 1967. It did not emerge as a viable methodology until 1972 when it was applied at the Kobe shipyards of Mitsubishi Heavy Industries [e.g., Hales, Lyman and Norman, 1990; Taguchi, 1987; and ASQC. 1992] in Japan. American Supplier Institute (AS!) and GOAL/QPC (Growth Opportunity Alliance of 1

Concurrent Function Deployment Chap. 1

2

Lawrence, Massachusetts/Quality Productivity Center) [e.g., Akao, 1990; and King, 1987] have done a great job in publicizing it in the United States. QFD was originally designed to take the voice of the customer (called customer objectives) and translate them into a set of design parameters that can be deployed vertically top-down through a serial four-phase process [Sullivan, 1988]. The four phases, known as ASI's four-phase process, are product planning, parts deployment, process planning, and production planning. The overall objective of QFD, which was quality deployment when introduced in 1967, today is still the product's quality. Emphasis on quality was also the reason why it was named Quality Function Deployment by the Japanese producers [Crosby, 1979; Deming, 1986; Taguchi and Clausing, 1990]. Recently Don Clausing and others have introduced some structural changes in the way the QFD information is arranged. The new arrangement is commonly called the extended House of Quality [Hales, Lyman and Norman, i 990; Taguchi and Clausing, 1990]. However, the original emphasis has not changed at all.

1.1

COMPONENTS OF QFD This extended house of quality (HOQ) consists of eight fundamental areas, all of which are not essential. Figure 2.23 of Volume I identifies the names of each area, and the door example in section 2.8.2.3 (see Figure 2.24) gives a glimpse of its full potential. In the following section, we visit each room of the extended HOQ and examine its essential features.

1.1.1

HOQ List-Vectors

Figure 1.1 identifies all rooms in the extended HOQ by its list-vectors and matrices. The four list-vectors-WHATs, HOWs, HOW-MUCHes, and WHYs-are described in the following sections.

1.1.1.1 WHATs: Customer Requirements (CRs) The customer defines what constitutes WHATs in a QFD/HOQ, In simple terms, this is a Hst of customers' wants. In most consumer goods manufacturing companies, Voice of the Customer (VOC) is considered the market requirement. Customers are initially listened to and a list of customer needs and expectations are created. This list is called WHATs or customer requirements (CRs). Some typical examples of WHATs might be "pleasing to the eyes," "looks well built," "provides good visibility," or "opens and closes easily." WHATs normally define a set of end-user requirements about what a consumer wants or likes to see in a future product. Dr. Noriaki Kano developed an expanded concept of quality. The Kano model of Quality or Features, as it is most frequently called, defines three types of WHATs: Basic, Performance, and Excitement. The Kano model relates customer satisfaction for each WHAT to its degree of achievement. An improved form of the Kano Model is shown in Figure 1.2. The degree of achievement is plotted against customer satisfaction. The two extremes of achievements are fully achieved and not yet started. The two extremes of

Sec. 1.1

Components of OFD

3

Key Product Characteristics (KPCs)

Customer Requirements (CRs)

Correlation Matrix

Product Characteristics Targets

Target Matrix

FIGURE 1.1

CRs Priority or Weighing Factors

Weighing Matrix

QFD Extended House of Quality-Ust~vectors and Matrices

customer satisfaction are ve1y satisfied, and very dissatisfied. This divides the domain spanned by the achievement and satisfaction axes into four quadrants, which in itself are divided into three regions. The spoken performance called performance WHATs constitute the middle region and are shown in Figure 1.2 by a flashlight- or a speaker-shaped boundary. The upper region of the flashlight model describes the areas for excitement WHATs. The lower region in the flashlight model shows the domain of unspoken or basic WHATs.

• Basic WHATs: These are the basic set of WHATs-the core quality feature or function set that a customer normally expects in a product. This is schematically shown by a region below the speaker-shaped area in Figure 1.2. It is assumed that customers do not have to ask for these quality features or functions, hence called unspoken. These features are typically hidden or implied functions of the products. Such features exist as a natural part or normal function of the product in some form, unless the product fails to work. The basic set of WHATs hardly ever increases customer satisfaction, but when left unattended, it can adversely affect customer satisfaction. One example of this basic quality feature or function is a car engine starting and not stalling. A customer would not necessarily be happy just because the car started or it did not stall, but surely enough, he or she will be frustrated and dissatisfied if it does not start or if it stalls frequently. Another example of a basic quality feature/function is that a customer would normally assume that the oil pump works and would not even think about it in a marketing survey unless, of course, it gives trouble.

Concurrent Function Deployment

4

Chap. 1

Very Satisfied Excitement WHATs

ThI.•,,

\

Gone ' Right

.

~r'

,,._:. -.. Perl'ormance

Excitement

t

WHAT•

0

Unspoken Excitement

m

e

• Not Started Yet

~

Fully Achieved

Basic o Expected or assumed o Typical of invisible product'>

s a

• oFunctionsoftheproducts o Natural part of the products

I I

r'

Things Gone Wrong

Unspoken Basic

; 1

- - - - -) -

a c

Unspoken Basic

i

Minimum Basic 0 . - - - - - - - - - - - - - - n --··--·-··-··-··-··-··-··-··--·-·--·-··-··-··----··--------·-·-----

Basic WIIATs

Very Dissatisfied Feasible

I Jegree of Difficulty

Simple

FIGURE 1.2 Three Types of WHATs (Customer Requirements)

Performance WHATs: They are usually determined through market research and are one-dimensional. These are often called spoken WHATs. A set of performance quality features describes how well a product measures up to the customer's wants. This is represented in the middle region of Figure 1.2 by a flashlight or a speakershaped area. Customer satisfaction increases with the degree of achievement. Examples of performance WHATs would be fuel economy, quietness, and a comfortable ride.

Sec. 1.1

Components of QFD

5

• Excitement WHATs: These are unspoken WHATs. They include new innovations and thoughtful engineering, providing pleasant surprises to the customer. These are shown in Figure 1.2 by a region above the speaker-shaped area. They increase customer satisfaction, even though the customer would not necessarily be dissatisfied without them. The 3M Corporation calls these WHATs customer delights. Examples of these are split sun visors, cargo nets in the trunk to hold down plastic grocery bags, holes drilled for fixing toddler seats, and so on. Over a period of time, a CR can shift character on the Flashlight model. This is shown in Figure 1.2 by a TIME arrow pointing down. Over time, a CR that was an excitement character at one point (a part of the upper half region) can take the form of a basic quality (a part of the lower half region) when products are matured. An example of this is power steering. This was an excitement CR in the 1960' s, a performance CR in the 1970' s, and now it is almost a basic CR. If a quality feature/function of a new product is contained in the upper half region, things are said to be going right for the product. The companies are expected to make a debut in the marketplace with such product introductions. On the other hand, if some quality features/functions fall below the speaker-shaped region, products are considered to be noncompetitive in the marketplace. The normal cycle of change is from a set of excitement quality to performance quality and then to a set of basic quality. The corresponding WHATs can further be categorized into primary (must have), secondary (may be), and tertiary (like-to-have) categories.

1.1 .1.2 HOWs: Quality Characteristics Items Manufacturers define what constitutes HOWs in a QFD/HOQ. This is represented by a list-vector in the Quality House marked as HOWs (see Figure L.3). In simple terms, HOWs are a set of quality characteristics through which a set of WHATs can be realized. HOWs thus represent an array of design variables or alternate solutions, which may or may not be independent. Each of the HOWs provides a solution or alternative for attacking one or more WHATs (or CRs). Manufacturers do not know what magnitude of each of these HOWs, when considered as a unit, will lead to the realization of as many WHATs as possible. HOWs provide an operational definition for the market quality characteristics. Using this list, a company can measure and control quality in order to ensure WHATs' satisfaction. The HOWs are the methods or techniques to translate the Voice of the Customer into design evaluation criteria. Typical entries on the HOWs list-vector are parameters for which means of measurements or a measurable target value can be established. For example, a customer need for a good ride (a WHAT) is achieved through dampening, shock isolation, anti-roll, or stability requirements {the four HOWs). The HOWs determine the set of alternate quality features to satisfy the stated customers' needs and expectations (WHATs). For this reason, HOWs are also called Quality Characteristics (QCs). A typical HOW might be a length, a width, a height, a thickness, a usable surface area, a volume, a set of material characteristics or mass properties, and so on. For every WHAT in the CRs list, there are usually at least one or more HOWs to describe possible means of achieving customer satisfaction. A best of the class product con-

6

Concurrent Function Deployment

Chap. 1

tains HOWs that satisfy a set of WHATs in some prioritized manner (see Figure 1.3). This is the chosen way in QFD of defining a relative priority for meeting the WHATs objectives. If a product solution (a HOW) exists today, a vector of such HOWs can be looked upon as strategic proportions in which customer requirements (WHAT.') are satisfied. A HOW is the way to assess feasibility of the product in the marketplace. A HOW list helps to define the target solution in relation to the WHATs list. 1.1.1.3 HOW-MUCHes: Bounds on Quality Characteristics This is a list-vector and normally identifies the bounds on the feasibility of HOWs. These entries are in the list-vector called HOW-MUCHes and represent the target values for each quality characteristic (see Figure 1.3). In other words, for each HOW on the list-vector, there is a corresponding value for a HOW-MUCH entry. The idea is to quantify the solution parameters into achievable ranges or specification table, thereby creating a criterion for assessing success. This information is often obtained through market evaluation and research. HOW-MUCHes capture the extremes-the permissible target values, positive or negative-depending upon theHOWs sentence construction or statements. A typical HOWMUCH measures the importance of HOWs, a performance of Product X, or a set of target

Design Trade-off

Design Characteristics

Current Market Definition Why this product need to exist? (List of Customer Groups, Competitors, Whats' Priority, etc.)

j

WHYs

=I

m

Customer Requirements

w

i=l

H A T



Relationships I =Weak or

None 3=Medium 9 =Strong

4

5 k-

2 I 6

k=p

Competitive Products

i=n I=!

Perceived Performance Sales Point Customer Importance Ratings Customer Competitive Assessment

HOW MUCHes 57 72 11 24 36 l =q

Technical Information Headings (Technical Importance Rating, etc.)

..,___ Overall Importance

Target Values (Maximum or Minimum)

Technical Importance Ratings

Technical Competitive Assessment

FIGURE 1.3 Expanded House of Quality-Terminology and Conventions

Sec. 1.1

Components of QFD

7

values. In an optimization formulation discussed in section 1.7, a row of HOW-MUCHes is used to collect upper and lower bounds for the attributes in the HOWs list-vector.

1.1.1.4 WHYs: Weighting Factors on WHATs Similar to WHATs and HOWs, a set of WHYs is also a list-vector that describes the relative importance of current competitive products, referred to as world-class or best of the class products. Once these important values are multiplied with the corresponding set of WHATs

and then su1nn1ed, they can provide a single pseudo measuren1ent index for overall customer satisfaction. In terms of optimization this can represent a weighted sum of objectives. An exa1nple of WHYs is a list-vector of relative importance with respect to customer wants for a world-class product of a major competitor. If the product is targeted to multiple customer groups, such as U.S., Asian, European, Japanese, and the like, this list must include these customer groups and their relative wants. WHYs are names of competitors, competitive products, market segments, or other items that describe the current market conditions. WHYs are also factors for weighing decisions that a future product must take into account. This usually translates into specifying weighting factors for WHAT,,. Setting priorities means specifying what is significant in the list of WHATs and what is not. A typical WHY might be a list-vector of overall importance, a vector list of importance to the world purchaser, or a set of world-class achievable performance of a product X. 1.1.2

HOO Relational Matrices

The four relational matrices are described in this section. HOQ Relational Matrices employ either numbers or symbols depending upon the purpose of QFD and the context in which QFD is being used (see Figure 1.3). Two possible rationales have been traditionally pro-

posed depending on whether a relational matrix is used for calcu1ations or for visual aid. Quantitative Reasoning: Numbers are used for specifying magnitudes of HOQ matrices. This facilitates comparing magnitudes of computed list-vectors through

mathematical means. • Qualitative Reasoning: Symbols are used to represent list-vectors or matrices. This provides a better visual communication. Three symbols are often used to indicate the relationship between the entries of WHAT,, and the HOWs. A solid circle (Gil) implies a strong rel;itionship, an open circle (0) a medium relationship, and a triangle (l!.) a weak or small relationship. This process of evaluating expressions in QFD gives a concu1rent team member a basic method of comparing the strengths and weaknesses, importance of column-vectors (WHATs, WHYs) or row-vectors (HOW:,, HOW-MUCHes), and measuring interactions between them. The notations used here follow the convention adopted by the employees at the Kobe shipyards who incorporated the local horse racing symbols. By convention, each symbol in the relationship matrix receives a value. Table 1.1 shows a convention that is typically followed in defining QFD relational matrices.

Concurrent Function Deployment

8 TABLE 1.1

Standard Relationship Conventions (Weight and Symbols)

Matrix WHATs versus HOWs

Quantitative

Grade

Strong relationship Moderate relationship Weak relationship None

HOlVs versus HOWs

Chap. 1

Grade Strong Positive relationship Medium Positive relationship Positive relationship None Negative relationship Medium Negative relationship Strong Negative relationship

Qualitative

Weight

Symbols

9 3 l

Double or Solid Circle (9) Open Circle (0) Open Triangle (6)

0

Blank

Weight

Sy111bols

9 3 l

Double or Solid Circle (e) Solid Triangle (A) +

0

Blank

-I

-3 -9

Open Triangle(~) Open Circle (0)

WHATs versus HOWs: Correlation Matrix Relationship between Market Requirements & Quality Characteristics (QCs) To get a relationship between market requirements and quality characteristics, a matrix is created by placing the HOWs list along the column of a matrix and the WHATs list along its rows (see Figure 1.3). The rectangular area between the rows and the columns then depicts the relationships between the set of WHATs and the HOWs. The matrix thus developed is called a relationship matrix. It correlates what customers want in a product and how an enterprise can achieve those objectives. The matrix-WHATs versus HOWs-is a core relational matrix of QFD. Relationships within this matrix are usually defined using a four level procedure: strong, medium, weak or none (see Table I.I). An example is shown in Figure 1.3. This matrix may be densely populated (more than one row or column affected). This results from the fact that some of the quality solutions may affect more than one market requirement. For example, what a customer wants in good ride and good handling (WHAT. Minimum or Maximum.

(l.27)

Subject to g}v)::;:: O; for j = 1,2, ... , m

(l.28)

V. ~v~V mm max

( I .29)

In most cases, F(v) is termed the objective function and g}v) the Constraints. Vmin and vmax denote lower and upper bounds on design variables. Note, the question we are posing is how can CFD be used in this setting. What elements of CFD represent design variables, constraints, and objec~ve function? Notice that

30

Concurrent Function Deployment

Chap. 1

rows of WHATs in CFD constitute customer wants or CRs. It is therefore similar to specifying a set of objective functions that a company must strive for, and pay attention to, in order to please the customers. Let us introduce some additional notation. WHATs s

l/;} for i = 1, ... , n

(1.30)

Clearly, all WHATs are not independent. It is also safe to assume that one or more WHATs can be satisfied individually by choosing the right HOWs. It will be very difficult, however, to satisfy all WHATs simultaneously by trial and error. Since HOWs are possible solution parameters that can be chosen to address WHATs, they are like perturbation variables of the original problem. In an optimization setting, vector HOWs of CFD will be like design variables. This means HOWss {v)

forj= l, ... ,m.

(1.31)

If HOWs represent system variables, in order to achieve an optimal solution, we must have these variables independent of each other. From CFD matrices we know that all HOWs are not independent. The correlation matrix, the roof of CFD denotes the dependencies of one HOW on another. In an optimization formulation we can pose such relationships between HOWs as constraints of the problems. This situation can be handled through what we normally call a linking technique in optimization. The correlation matrix gives a linking matrix for the design variables. This matrix can be formulated as follows: {v,} =[HR] {v,l

(1.32)

where HH is a square matrix of size (m, m), or

v={HH)Txv s r.1 r

(1.33)

where HH is a triangular square matrix whose diagonal tenns are all zero. This is because in the CFD correlation matrix for HOWs versus HOWs, H,s=O"lfr=s,

(1.34)

where rands are the rows and columns of the HOWs matrix. The above relation simply says that if we consider HOWs as a set of design vectors, all of the terms in the HOWs vector are not independent of each other. Equation 1.32 or 1.33, thus, serves as a linear set of equality constraints to the corresponding optimization formulation. Let us define a term of HM matrix as hm1r Since, in equation 1.31, we denote a row vector of HOWs as {v}, by definition of a HOWs versus HOW-MUCHes matrix, a value stored in hm1j would represent a limit on v; If we denote two vectors of the HM matrix (I= Land U) representing permissible upper and lower bounds on HOWs, the following statements can be made, hmLj 5, vj 5, hmuj

forj = I, ... , m

(1.35)

where hmL denotes a lower bound on v. and hmu. denotes an upper bound. 1 It is 'interesting to note that equ.{iion 1.35 represents an equation of bounds on the system variable for an equivalent optimization formulation.

Sec. 1. 7

Formulation of CFD as an Optimization Problem

31

So far we have introduced the concept of design vectors. bounds on the design variables, and a series of objective functions defined by a WHATs vector. The only tenn left in completing an optimization formulation is a weighted objective function. Earlier we also noted that a list of WHATs represents a series of objective functions. We also noted from the matrix WW that a column of this matrix contains the weighting factors associated with CRs, the WHATs. If a pth column of WW matrix contains this factor, ww. , a cumulative or weighted function, F(v), can be fanned using TIR formulation (equatiori' !.26) as follows: FPl.(v)= [[wh.]'x {ww )] I] 1p

(l.36)

F .(v)=[{wh).T {ww) ].

(1.37)

or in a matrix form, Pl

P

J

Note that in Equation 1.36 or 1.37 a weighted multi-objective function has been formed that is the weighted sum of all its components. The result can be stored as a qth row of an HM matrix. wherej= I, ... , m.

That is,

(1.38)

In WHATs versus WHYs, sub-section I .1.2.2, many ways of obtaining tl1ese weighting factors were described. Equation 1.38 represents one of many cumulative fonns that are possible. Now, if F(v) is employed as an objective function the general statement of optimization previously described in section 1.7.1 can be transfonned in CFD context as described in the following section.

1.7.2

CFO based Optimization

The original optimization problem as stated in section J.7.1 can now be transformed. Find a vector of design variables v E D, where {v) = HOWs, that minimizes the objective functions Fr/v),

where or

F{J).(v)=[[wh.,Yx {ww.) l · I) I pJ {F(v)) p = [WH]' {ww) p

(1.39)

(l.40)

where i = 1, ... , n; j = 1, ... , rn; and p is a column index of a WW matrix, while satisfying the following constraint equations

{v,J = [HH] ( v,}

(l.41)

where rand s range from 1 to m. The bounds on the design variables can be expressed as: hn11.J $'

or

{hmL)

vj

$

hniVj

:> {v) :> {hmul·

(1.42)

(l.43)

In equations 1.39 to 1.43, the vectors {wh) and [ ww} belong to CFD/HOV relational matrices, WH and WW respectively. Lower and upper bound vectors {hmL) and {hmu) are the rows of the CFD/HOV relational matrix, HM.

Concurrent Function Deployment Chap. 1

32

1.8 HORIZONTAL DEPLOYMENT Jn sections 1.5.3, a basic case of CFO for quality FD was described. Similar to a transformation of a Quality FD from a qualitative approach to rigorously applied quantitative methodologies, each of the TVM life-cycle values can also be transformed following this approach. Figure 1.5 shows the 6 artifact values (AVs) along a horizontal x-axis and five value characteristics (VCs) along a y-axis. The z-axis shows the vertical deployment process of each of these (x-y) combination groups through its three tiers: the product planning, process planning, and production planning tiers. The vertical arrows (z-axis) show the sequence of steps during the tier deployment and the horizontal arrows (x-, and y-axis) show a process of concurrent deployment. A typical list of WHATs for Quality FD values is shown in the next section. In the following sections, WHATs have been identified along with a set of VCs. This is useful if a separate CFO matrix is required for each artifact value. If there are not many items for VCs, some of the items can be combined. One of the goals of CFO is to develop products flexible enough so that many life cycle TVM perspectives may be run in parallel without violating or contradicting any one of the previously assigned WHATs. The optimum result would be derived from deploying a great number of TVM values simultaneously, some of which-new materials, cost estimates, process technology-may not always be available.

1.8.1

First Level Deployment Metrics of Artifact TVM Values

In the following sections, the first tier Function Deployment (FD) metrics are defined for each TVM value. For the initial tier, a generic set of WHAT.' (requirements) and HOWs (characteristics) are identified. Besides the two list-vectors, future concurrent FD metrics must include quantifiable measures that can be used directly in performing values' tradeoff analysis. Only with the set of quantifiable measures, would the product performance designer, cost estimator, X-ability checker, responsiveness keeper, and infrastructure manager be able to adequately evaluate an artifact's conformance to TVM requirements.

1.8.1.1 Quality Deployment Quality is the first artifact value that needs to be deployed. Products designed without built-in quality during design and quality assessments after production-relea•e suffer from a number of shortcomings. Quality problems translate into higher manufacturing costs, particularly when some defects get out to the customers. The two list vectors, WHATs and HOWs for a HOQ matrix, are listed in Table 1.2. 1.8.1.2 X-ability Deployment Most companies set goals on a limited set of X-ability considerations. Product performance and quality are generally considered the basic needs. They are separately specified and explicitly dealt with. Besides the major key ones, such as functionality, reliability and maintainability, there are many other factors that need to be quantified. These are parts

Sec. 1.8 Horizontal Deployment TABLE 1.2

33

Examples of WHATsand HOWsin a Quality Deployment

Examples of WHATs Component packaging Concept design Design criteria Detail design DifficulHo-handle parts Experimental testing Material handling Material properties Customer satisfaction Nonstandard routing Pre-production Process planning Product shape Production Prololype On-time shipment Validation Warehousing

Examples of HOWs Managing manufacturing precision or quality DFM guidelines Computer Aided Design (CAD) Design axioms Managing detailed dimensions, roundness, etc. Just-in-Time (JlT) Continuous improvements Ensure robustness Process description rules Specify minimal eccentricity Formal design reviews Informal design reviews Minimum parts assessments Material and process selection Assembly evaluation method Managing tolerances Quality Control (QC) Good surface finish. texture. etc.

count simplicity, parts joining efficiency, mating parts assemblability, sheet metal parts formability, plastic parts producibility, and so on. Tables 1.3 and 1.4 outline a list of such WHATs and HOWs belonging to this X-ability class. The problem is really in deciding what is important about an aspect of design and expressing the corresponding value X-ability (HOWs) attributes in a way that addresses the customer objectives.

1.8.1.3 Tools and Technology Deployment In Section 2.8 of volume I a number of life-cycle tools were outlined. They were categorized as computer tools and technology tools. (See equations 2.86 and 2.87 of volume I). The impact of a tool deployment on the resulting organization is not the same. Table l.5 outlines a list of WHATs and HOWs belonging to this tools and technology class. 1.8.1.4 Cost Deployment ln Chapter 2, volume 2, of this book, Design for X-ability (DFX) is described. The techniques employed during DFX methodologies, are either handbook type procedures, or a collection of rules and guidelines (heuristics) of what.might be considered the best design practice. The aim of such DFX techniques is usually to meet the provision of X-ability guidance one at a time. ln concurrent engineering, the problem of collectively evaluating a design is more fundamental than provisions for judging the designs one at a time. DFA, like many another Design for X-ability, represents only a single consideration for obtaining a better design. In DFA, when two separate components are integrated into one, the latter is considered better because it ensures easier assembly than the old ones. Often such recommendations lack quantification. The designer is usually not told how the new design is better

34 TABLE 1.3

Concurrent Function Deployment

Chap. 1

Examples of WHATs in an X-ability Deployment

Examples ofWHATs

Brief Descriptions

Simplicity

Simplicity is perhaps the most significant attribute of a successful design. Simplicity signifies many things to many persons. It may mean fewer parts, parts that perform multiple functions, parts that are easier to assemble (manually or by robots), or parts that can be made by simple processes. Most engineers and designers are intrigued by the complexity of parts. Very few really know how to design for simplicity.

Reliability

Manual product assembly may suffer from dimensional deficiency (e.g., Jack of adequate fits and clearances), variability (e.g., precision machining) and non-uniformity in assembly (e.g., lack of adequate accuracy and consistency), etc. They ought to be minimized. Impact of equipment reliability on process needs ought to be minimized. Designers must consider manual options in all facets of production as stand-by alternatives in the case of castrophic failures.

Maintainability

Normal wear and tear, unexpected malfunctions, and alike necessitate maintenance. To increase product reliability, maintenance considerations must include a checklist for troubleshooting, such as how to bring the fault equipment quickly on-line, and normal inspection as standard DFM procedures. Total preventive maintenance is usually considered a part of design for 1naintainability.

Serviceability

Serviceability includes ite1ns such as: Repair parts Testability Inspectability

Parts Consolidation

Besides many other benefits, parts consolidation reduces the need for using different materials for various parts of a product. It also increases the value oi a single product to the collector of recycled parts. The receiver of the recyclable parts will not have to decipher each part to recover materials. • Material Selection Strategy: With high volume products, broader use of the same material and with low volume products, market value of the higher end material (such as resins), make disassembly and recycling more attractive. Ease of assen1bly: Parts consolidation or incorporation of multiple parts into a single part reduces the nmnber of materials to be processed in the assembly and recycling process. It also focuses on elin1inating unnecessary assembly operations since these parts need not be assembled. • Ease of Disassembly: Parts afler consolidation should not be made so complex that the process of disassembly becomes difficult and time-consu1ning. If that occurs consolidation is counter-productive.

Assembly/Disassembly

Fastening methods (employed during assembly) dictates the level of difficulties during disassembly. There are two ways to disassemble a part, by reverse assembly or by brute force. The following are various methods used: • Adhesives • Heat Staking • Induction \Velding • Inserts Screws • Snap-fit Latches • Thermal Methods

ProducibilUy

Material selection strategy and its appropriateness impact significantly the cost and part performance and also influence the selection of manufacturing processes. Process classification is dictated by the approach taken for product manufacturing such as net shape or near net shape machining. Near net shapes require secondary operations of finer finish and tolerances.

Sec. 1.8 Horizontal Deployment Exa1nples of WHATs

35 Brief Descriptions

Manufacturing process selection should be made after detailed evaluation of a11 economic and competing life-cycle factors. Design geo1netry features establish the shape of the part. Production quantity such as minitnum production volu1nes is a function of the required level of product quality, such as tolerances and surface finish.

Product Usability

Product usability emphasizes making the work group members the focal point of the design, stressing ease of use including skill requirements for learning how to operate. Products should be designed to fulfill work group needs.

Availability

The work groups should have access to design histories, old CAD files. design database, technical memory and library of parametric parts, etc.

Supportability

This includes both the personnel support and the virtual support extended by the network of computers (both hardware and software).

bistal/ability

The virtual library of software programs should be installed on all hardware platforms, firmware, and operating syste1ns, etc. and work seamlessly as if ir was developed for that platfonn. The new version of all electronic computer programs should be upwards compatible. The old data files or objects should be able to work well with new programs or new objects without problems.

Upgradeability

Disposability

If materials are health hazards or inflam1nable, work groups should find ways to devise a waste or fire management system for their safe disposal.

Recyclability

When large parts use the same materials, or when less disassembly is required to recover the recycled inaterials (such as plastics), the recycler is guaranteed a better opportunity to recover investments through recycling.

TABLE 1.4

Examples of HOWs for X-ability Deployment

Examples of HOWs Assembly and Syste1n Considerations Design Criteria Engineering Change Order/Note {ECO/ECN)

Brief Descriptions 111is relates to design for assembly or syste1n design using consistency of purpose. Design criteria are considered a set of constraints to be satisfied. Design in its life cycle are subject to a number of ECOs and

ECNs. Features and Tolerance Paris Classification

Quality Function Deployment (QFD) Quality Inspection and Tooling Risk Manage1nent Standardization Syste1ns Engineering Value Engineering

Cenain features and tolerance results in better design or bet~er representation for the design capture. Classification of parts helps in concurrency and downstream operations such as GT-based process planning, etc. QFD stream sorts the relevant requirements fro1n those that are redundant or unnecessary. Qualiry inspection and tooling is a combination of errormonitoring, measure1nents, and feedback. This is described in section 2.2.1 of volume I. This is discussed in Chapter 5, section 5.8 of volume II. This is discussed in Chapter 6 of volume I. This is discussed in section 3.6.2 of volume II.

TABLE 1.5

Examples of WHATs in Tools and Technology Deployment

Examples of lVHATs

Workstation

Information Managen1ent

Teain-based Requirements

Brief Descriptions This includes bringing in MWS power to reduce the computational burden and wait time. Design and operator training should account for variations in population, skill levels, and other human stamina and con1fort needs. Information r..1anagement consists of two parts: inform
Key Produmon Coatrol (Poe)

...,

"""""1m ......

Target M.u-b: ltl.1

Wcigbillg

\

\

"""" ,_ !IL2

..

P'1C Priority «Weighing

Produc&n Piannilfgfor Sub-iystem

Linked House of Quality-Deployment of RCs Attributes for Quality along the

z~axis

46

Concurrent Function Deployment

Chap. 1

Voice of

Customers

c c

c

Mointennace

c

Verification

c

J

Assembly/Build

Subsystem

c

J

Components

)

J

(Parts Characteristics)

)

c

(Education/Trainin~

c

)

Sys!=

Field Support

Materials

)

Assembly (Process Planning

J

c~_s_~_ing__)

(Design/Build Planning)

C~_Machinin_··-'s_J

FIGURE 1.14 Planning Operations of CFD (z·axis Deployment)

requirements through each tier of product, process, and production planning levels. The basic approach CFD uses is conceptually similar to the step by step product development approach used in the CE process discussed in Chapter 9 of volume I.

1.9.3 Typical Planning Parameters A typical set of generic infonnation that is required during vertical deployment (three tiers of a CFD) is defined in Table 1.12.

Sec. 1.10

47

Implementation Issues

oCost

oTime

o Features o Economical use ofresources o Customer satisfaction

High Risk

To next sub-phase or new phase level

FIGURE 1.15 Filtering Method in CFO to Narrow Down the List of Available Options

1.10 IMPLEMENTATION ISSUES Accurate creation of the CFD house of value (HOV) vectors and matrices is not a simple task. The data has to be consistent and should represent real world situations. Implementation requires putting together a team of experts familiar with the products, processes and the con1petition. Inputs are required from various sources including inanagement counci1.

engineering steering committee, quality control directors, n1arketing product managers, manufacturing steering com1nittee and strategic business units. It takes a long time to come up with substantiated data that everyone is comfo11able with. Each must understand the mission of the product introduction, its goals, strategy, and key success factors. Assumptions and ground rules should be laid down to guide each team member into the

same goals and directives. In this chapter, a framework for deployment is described. The CFD methodology exploits the independence of units that manifest itself in Strategic Business Units, TQM, and Total Enterprise Management concepts that are now emerging. It considers parallel deployment, as contrasted to the phased deployment, for example, American Supplier Institute's QFD concept [Sullivan, 1988], in meeting life cycle values. It is not based on using a single measurement, such as quality as in QFD. The present approach is more versatile than Akao' s GOAL/QPC concept. In the present setting, AS!' s QFD emerges as a special case of concurrent function deployment. It enables the Planners and Strategic Decision Makers to deal early with tradeoffs among the crucial factors of artifact values. Six concurrent values, namely functionality (Quality), perfotmance (X-ability), tools and technology (innovation), cost, responsiveness, and infrastructure (delivery) are considered simultaneously rather than serially.Three-dimensional Value Characteristics Matrices (VCM) are employed to ensure that both the company and tl1e customers' goals are optimally met. The key artifact values are

Concurrent Function Deployment Chap. 1

48 TABLE 1.12

Planning Information Required during Vertical Deployment Tiers

Examples of Product Planning

Voice of the Custo1ner: gathering requirements, interview, survey, etc. Marketing: needs of market, customers, collection of technical data, information Planning: product strategy, long term technology development, sales plan, etc.

Research and Development: develop ideas and off-the-shelf technologies. Develop: development target, critical control, technical development

plan Design: making of part and audit of design Analysis: Failure mode and effects analysis (FEMA), fault tree analysis (FTA), etc. Engineerillg: Engineering the product functions Prototyping: prototype design. prototype order, etc.

Testing: prototype testing, evaluation, etc.

Design standards: methods and 3Ps, features, and parameters

Examples of Process Planning

Examples of Production Planning

Product characteristics: determination of go/no go for design Process audit hnprovemcnts

Process characteristics

Network of assurances

JIT; daily production control

Part Classification; group technology

Purchasing/outsourcing

Plant Designation: the person or the work group in charge of production Processing/equipment

Manufacturing and Assembly

Mass production

Pre-manufacturing: QA charts, mistake proofing

QC process charts, parts inspections, shipping inspections

Process Analysis

Sales

Process Design: Process Deploy1nent, control plan, NC, Machining, grinding, preproduction prototype Pre-production evaluation performance test, process capability study Process standards: 1nethods & procedures, process continuous improvements

Claims Handling

Production audit improvemetlts

Maintenance and service, delivery and support Production standards: methods & procedures, production continuous in1prove1nents

deployed in parallel tracks, stressing their importance and making it less likely to have them ignored by omission. CFD is a customer driven process for concurrent product design and development. CFD is a system of identifying and prioritizing customers' needs obtained from every available source. By using CFD, engineers can methodically analyze the details of design and process improvement to meet those needs. CFD is a highly versatile engineering tool translating the Voice of the Customer through all phases of the product life cycle. In this chapter we have shown that the concept of CFD can be used to develop a world-class product in the least possible duration (elapsed time). The key feature of this approach is that it is based on generally accepted QFD practices, conventions, and nomenclatures. It uses the same set of list-vectors and relationship matrices that were previously defined.

49

References

The concept, however, is much 1nore useful and versatile than conventional deployment as in QFD. The trio techniques allow the users to optimize concurrently all sorts of product characteristics under a unified set of WHATs that may belong to a number of tracks, such as X-ability, cost, responsiveness, too1s and technology, and infrastructure, among others.

REFERENCES AKAO, Y.A. 1990. Quality Function Deployn1ent-lntegrating C11sto1ner Requiren1e11ts into Product Design. Cambridge. MA: Productivity Press Inc. Also in Qua/it)' Deploy1nent, a series of articles edited by Yoji Akao, Japanese Standards Association, translated by Glen Mazur. Methuen, ~1A: GOAL/QPC.

ASQC (American Society for Quality Control). 1992. Automotive Division, American Supplier Institute (AS!), GOAL/QPC, 1992. Transactions frorn the fourth syrnposiun1 on Quality Function Deployntent (QFD), June 15-16, Novi, MI. ASWAD, A. 1989. "Quality Function Deployment: A Tool or a Philosophy." SAE Paper No. 890163. Society of Auto1notive Engineers, International Congress and Exposition, Feb1uary 27-March 3, 1989. W.R. 1992. Tools for Today's Engineer-Strategy for Achie11i11g Engineering Exceltence: Section 1, Quality Function Deploy1nent, SP-913, SAE Paper No. 920040. Proceedings of the SAE International Congress and Exposition, February 24-28, Detroit, ML

CAREY,

CLAUSING, D. 1994. Total Quality Develop111ent: A Step-by-Step Guide to World-l'lass Concurrent Engineering. Ne\V York: ASME Press. CROSBY, P.B. 1979. Quality is Free: The Art of Making Quality Certain. New York: McGraw Hill. DEMING, W.E. 1986. Out of Crisis, 2d ed. Cambridge MA: MIT Center for Advanced Engineering Study. DIKA, R.J. and R.L. BEGLEY. 1991. "Concept Development Through Team\vork-Working for Quality, Cost, Weight and Investment.'' SAE Paper No. 910212, pp. 1-12. Proceedings of the SAE International Congress and Exposition, February 25-March I, Detroit, MI: SAE.

CRPII (beyond QFD)." SME Paper No. MS90-03, Mid-America '90 Manufacturing Conference, April 30May 3, 1990, Detroit, Ml: SME.

FREEZE, D.E. and H.B. AARON. 1990. "Customer Requirements Planning Process

R., D. LYMAN and R. NORMAN. 1990. "Quality Function Deployinent and the Expanded House of Quality." Technical Report, pp. 1-12. OH: International TechneGroup Inc.

HALES,

HAUSER, J.R., and D.

CLAUSING.

1988. "The House of Quality." Harvard Business Review,

(May-June) pp. 63-73. Volume 66, No. 3. KING, B. 1987. Better Designs in Half the Tbne-lmplenienting QFD Quality F11nclion Deployn1ent in America. Methnen, MA: GOAL/QPC.

KROLL, E. 1992. "Towards Using Cost Estimates to Guide Concurrent Design Processes." PED¥Vol. 59, pp. 281-293. Concurrent Engineering, ASME, edited by Dutta, Woo, Chandrashekhar, Bailey, and Allen, Proceedings of the Winter Annual Meeting of ASME, November 8-13, 1992, Anaheim, CA: ASME Press. MIZUNO, and Y. AKAO. 1978. (ed.): Quality Function Deployment, JUSE (published in Japanese).

Concurrent Function Deployment

50

Chap. 1

PRASAD, B. 1993. "Product Planning Optimization using Quality Function Deployment." Chapter 5 In Al in Optiff1al Design and Manufacturing, edited by Z. Dong, and series Editor Mo. Jamshidi, Englewood, NJ: Prentice Hall, 117-152. PRASAD, B., R.S. MORENC, and R.M. RANGAN. 1993. '"Information Management for Concurrent Engineering: Research Issues." Concurrent Engineering: Research & Applications, Volume I, No. 1 (March 1993). QFD/CAPTURE. 1990. Users Manual. International TechneGroup Inc., Version 2.2. SULLIVAN. L.P. 1988 ...Quality Function Deployment." Quality Pragress, Volume 21, No. 6 (June). TAGUCHI, G., and D. CLAUSING. 1990. "Robust Quality." Han1ard Business Revie\v, Volume 68, No. I, pp. 65-75, (Jan.-Feb.). TAGUCHI, S. 1987. "Taguchi Methods and QFD." Hoivs and Whys for Manage1nent, American Supplier Institute, Dearborn, MI: A.SJ. Press.

TEST PROBLEMS-CONCURRENT FUNCTION DEPLOYMENT 1.1. In QFD, there are 4 phases that deploy Voice of the Customer (VOC) to get to an improved product. What are the components of QFD? Explain each of the four QFD phases and give examples.

1.2. Besides QFD, Taguchi and Pugh concept selection, can you think of other engineering tools that can help in concurrent engineering? What prevents a team from not using them during an early product introduction stage? 1.3. What makes QFD powerful? How can one use Value Engineering with QFD? Show a flowchart of the two working together. 1.4. When is QFD not useful? Many of the HO\.V.s (so called quality characteristics) in QFD are not all independent. How is this condition handled in QFD? 1.5. When customers want to buy a product, what qualities (values) of a product influence them to purchase it? What is voice of the customer (VOC)? When do they differ from so called customer requireinents? 1.6. How can the Kano Model be used to prioritize a set of custo111er requirements (CRs)? How does a CR shift character? When does that happen? 1.7. What are the rooms of HOQ? Why are Technical Importance Ratings (TIRs) listed under a HOW-MUCH list-vector? 1.8. What is the significance of weighting factors in co1nputing TIRs? How can manufacturers use TIRs to prioritize the quality characteristics of a product yet to be launched? 1.9. What are the limitations of deploying a QFD? What is required in optimizing an artifact to be recognized as the best in every class? 1.10. What are the pitfalls of a conventional QFD? Why is quality deployment not enough to lead a company to be a world-class product manufacturer or a service provider? 1.11. What n1akes a great product? Can deployn1ents of many value functions proceed in parallel? What will be their impact on the time-to-market aspect?

Test Problems-Concurrent Function Deployment

51

1.12. What is CFD? How dOes it differ fro1n QFD? How does CFO support long-range thinking and better commuriication across several value functions, Vi.'ork groups, etc.'! 1.13. Describe a CPD architecture. How does it differ fo1n a QFD architecture? \Vhat are the three main dimensions of a CPD deploy1nent?

1.14. Why is CFO concept with regard to quali(v FD equivalent to a QFD concept? Explain a trio deployment methodology for CFD and explain how it works. 1.15. CFD results into a three-dimensional house of values (HOV). Describe its major roo1ns. Where is HOV different from a house of quality (HOQ)?

1.16. What benefits does CFD presenr during quality deployment that are not evident during standard QFD process?

1.17. Show an examj:>le of a linked CFD House of Values for quality during vertical deployinent (z-axis, as shown in Figure 1.9)

1.18. Show an example of a linked CFD House of Values for quality during axial deployment (yaxis, as shown in Figure 1.10).

1.19. How would you deploy cost functions for design and development of a 100-pin stapler? 1.20. How \vould you deploy X-ability functions for design and developn1ent of a multi-color ball point pen?

1.21. How would you deploy tools and technology functions for a cooperative CE work groups in a CE-based organization

1.22. How \vould you deploy infrastructure functions for the cooperative CE work groups in a CEbased organization'? 1.23. How would CFD look for a service organization? What service values would you consider for a CAD/CAM or CAE service organization? 1.24. Describe the significance of filtering in QFD? What other dimensions would filtering take when applied to CFD?

1.25. How \vould you use CFD in formulating an optimization-based design proble1n? 1.26. How can CFO be used for company product assessment? Outline two rating schemes for this assessn1ent?

1.27. List five major advantages of CFD that can be used for the life-cycle management of a product. What should manufacturers do to become a world-class product manufacturer?

1.28. How can CFD be used for competitive product assessment? Outline two rating schemes for this assess1nent?

1.29. What are the implementation issues of CFD? 1.30. If CFD uses the same set of QFD practices, conventions, and nomenclatures, why is it a better concept compared to QFD?